[Amps] DI water explaination TSPA

[John T. M. Lyles]

Paul (below) is correct that Altronics loads (and 
Bird) are thin film resistive elements on a core, 
with water flushing by them. There is not 
requirement for super low conductance water in 
them, as the RF voltage is across the hot side of 
the coax to ground, and is shunted by a 50 ohm R. 
Guess where the RF current will flow.?

Bringing cooling water across HV requires a 
different water. DI is the best and is 
recommended by tube manufacturers, although it IS 
corrosive. But there is a difference in the 
metal-scavenging power of the water depending on 
its temperature, ph, and resistivity. Take a look 
at the Eimac paper on cooling water, I forget the 
title, but they said it all. RCA/Burle also has 
similar advice in their papers. Ditto, E2V, 
Thales, etc.

When we build high power systems that require 
cooling, and the water crosses HV gradient, the 
hose diameter, the length of the hose, the type 
of piping, and so forth are all critical for the 
proper design. I had assumed when I said to use 
DI water, that all of this was accounted for by 
the amplifier maker (Alpha?).

You don't use DI water with steel or iron piping. 
Use stainless steel, brass, copper, or plastics. 
Do not allow the resistivity to get up to maximum 
(18 Mohm-cm) as it will become very corrosive to 
metals, including the tube cooling jacket. That 
is why it is good to try and keep it in the few 
Mohm-cm if you can, and make the hoses long 
enough, or the diameter small enough, that the 
'resistance' of the water hoses (2 of them in 
parallel remember) prevents appreciable current 
from flowing. How much is this? Depends on your 
system. For big tubes, a few Ma may be 
acceptable, as the amount of metal which m 
migrates is small enough that damage is minimal 
for the lifetime of the tube (few years). In a 
small system, it should probably be uA. The 
larger liquid and vapor cooled (and multiphase or 
hypervapotron cooled) tubes will have a 
sacrificial metal anode (or is it cathode?) in 
the water hose fittings at each end of each hose. 
With this, the plug will slowly burn away, 
leaving ions in your system. It is inspected 
regularly, at least yearly, to determine the rate 
it goes away. When it reaches the edge of the 
metallic fitting, then it is replaced. This keeps 
the DI water from taking away important metal 
(like tube fittings, pipe fitting at ground, 
etc.). This is only needed across the HV 
boundary, not throughout the water circuit. One 
more thing, oxygen is a bad thing to have in a DI 
water system, so it needs to be removed (with 
another Culligan style bottle).

The idea of hipot testing water can be done, if 
you use a nonconductive pipe and want to measure 
the leakage in a known volume of DI water. This 
won't work for tap water, hence the need for DI. 
I hipotted a tube in circuit last week, with 5 
M-ohm-cm DI water. I had about 0.5 mA of leakage 
in the hoses, so the high potter must be able to 
handle that without tripping off. It worked fine, 
and I discovered a failed plate blocking 
capacitor without removing the tube. Not 
impossible, nope just did this.

I have never used anything but DI water and most 
other high power RF systems for acclerators and 
for broadcast (Continental Electronics in Dallas 
makes many) use the same. But one must be very 
careful of the fittings, the resistivity, etc. 
Probably beyond the scope of what we are talking 
about for ham use in an amplifier, perhaps. There 
are some excellent papers on this, from Argonne 
National Lab, which I can refer folks to if they 
are interested further.

John

At 2:32 PM -0500 2/6/07, amps-request at contesting.com wrote:
>Message: 7
>Date: Tue, 6 Feb 2007 14:16:14 -0500
>From: "Harold  Mandel" <ka1xo at juno.com>
>Subject: Re: [Amps] WATER for vapor cooled amplifier
>To: "'Robert B. Bonner'" <rbonner at qro.com>, <fsiyfr at okeechobee.com>,
>	<amps at contesting.com>
>Message-ID: <006001c74a23$47081010$630ca8c0 at laptopHM>
>Content-Type: text/plain;	charset="us-ascii"
>
>Okay, so let's back up one step and sorta tabulate our findings
>and observations:
>
>1. Someone found that Deionized water was too conductive.
>2. Altronics says never to use DI water in loads.
>3. Someone found distilled water too conductive.
>4. Altronics says distilled water is too reactive, also.
>5. Some people find Zephyr Hills distilled water
>     not too conductive and okay.
>6. Altronics says clean, potable tap-water with less
>     than 500 parts per million of debris is okay for
>     RF loads.
>7. BIRD says distilled water for loads is best.
>8. Paul found one water source supposedly good, a
>     dead short.
>
>My head is beginning to spin!
>
>The idea of putting the HiPot tester in the water seems
>the top winner today.
>
>Water, at it's best is supposed to be at pH 7.00, yes? 
>Then, it's at 1.00 Specific Gravity, yes?
>That should mean it has a certain conductivity, yes?
>
>Why would potable water, not distilled or DI, have a different
>conductivity between brands?
>
>Discounting the DI story, why would distilled water be
>harmful? Why would it have differing conductivities
>from brand to brand?
>
>Lastly, Altronics says to include a "people-safe" antifungal
>agent in recirculating systems that does not include salts
>or any other caustic agents. Do any of you use antifungal
>materials in your recirculating or vapor-phase cooling rigs?
>
>Hal
>
>
>
>
>------------------------------
>
>Message: 8
>Date: Tue, 6 Feb 2007 14:32:29 -0500
>From: "Paul Christensen" <w9ac at arrl.net>
>Subject: Re: [Amps] WATER for vapor cooled amplifier
>To: "Harold  Mandel" <ka1xo at juno.com>, "AMPS" <amps at contesting.com>
>Message-ID: <030f01c74a25$8a802dd0$f9153405 at OfficeDell>
>Content-Type: text/plain; format=flowed; charset="iso-8859-1";
>	reply-type=original
>
>>  "Okay, so let's back up one step and sorta tabulate our findings
>>  and observations..."
>
>A significant difference in the use of water with RF loads is that it is not
>mixed with high-voltage and accordingly, probably no real reason to minimize
>leakage, but immunity to long-term corrosion is important.  Typically, in a
>water-cooled load, water flows through a glass jacket which, surrounds a
>50-ohm resistive element.
>
>In my prior life, I used a 25KW Altronic water-cooled load on a Harris
>FM-20K commercial broadcast transmitter.  We simply ran a hose from the 46th
>floor water tap and it remained that way for 15+ years without encountering
>any trouble.  In all fairness though, I believe the building (downtown
>office complex) used a soft water system.
>
>On the other hand, since the vapor-cooled amps require good HV isolation, it
>would seem that the water requirements are much more restrictive with two
>overarching requirements: (i) low water conductivity; and (ii) low corrosive
>action.  I am not sure what the right answer is but the Zephyhills product
>labeled as "distilled" seems to work well with my Alpha 70V.
>
>Paul, W9AC